The previous section has outlined the effects of radioactivity and the limits placed by the ICRP and AEC on radioactive releases from nuclear power plants. The AEC regulations are based upon maximum exposures during or following a release at points on specified site boundaries, which are also part of the regulations.

5.2.1 Radiation Safety

One of the barriers mentioned at the commencement of this chapter was the physical distance which separates the public from the nuclear power plant. This barrier is defined by the following boundaries:

+ Higher altitude changes average 300 ft/5 mrem/yr increase.

a. Exclusion area. This is the area over which the reactor licensee has complete control. Residency is prohibited, or if there are residents, they are subject to immediate evacuation. There can be roads and waterways, but arrangements are made to control traffic if the need ever arises (Fig. 5.3).

b. Low population zone. This is the area around the plant that has a sufficiently low population so that arrangements can be made for their health and safety by evacuation or shelter. An exact population is not specified by the regulations, although the AEC would review the arrange­ments made for the emergency protection of residents in the zone (Fig. 5.3).

c. Population center distance. A population center containing more than 25,000 residents is considered a densely populated area, therefore the plant should be no nearer to such a population center than four-thirds of the low population boundary distance (Fig. 5.3).

It is clear that these boundary limits will vary for different plants in different areas. At the present time, for the LWR systems, the exclusion distance varies between 0.2 and 0.5 miles, a normal site boundary fence distance, while the low population zone varies between 1.4 and 7.4 miles. The radiation dose limits for these boundaries are set out in full in Section 5.1.3.3.

5.2.2 Other Siting Considerations1′

Many other practical considerations determine the choice of a site of a fast reactor power plant besides the radiation limitations. The following list enumerates some siting characteristics of interest.

(a) Population density. Is sufficient unpopulated area available to meet the requirements of the previous section?

(b) Present uses of the area. Could a power plant be built in the area without disturbance to the present use of the area; for example, is it all prime agricultural or conservation land?

(c) Physical characteristics (geology, seismology, hydrology, and meteo­rology). Would the physical characteristics of the site make a power plant overcostly or unlicensable in view of the geological, seismological, hydro­logical, or meteorological difficulties?

(d) Economic location with respect to the load center. Is the site close to or far from the center of load demand or a suitable distribution grid?

(e) Proximity to water or convenience of cooling ponds or towers. Can sufficient and adequately reliable cooling water be supplied (see also Sec­tion 6.4)?

(f) Site size. Is the site large enough for any future expansion planned?

(g) Transportation facilities. Are adequate roads, railways, and water­ways available for incoming plant components during construction?

(h) Labor availability. Is a labor force available in the proximity of the site to provide construction labor?

(i) Attitude of the local community. Is the local community liable to help or hinder the project? What local political situation might affect the site (see also Section 6.4)?

(j) Reactor characteristics. Does the site have any particular attractive­ness for the type of reactor under consideration?

These and other similar questions all require answering before a site can be selected. Many of these questions are not safety matters, but almost all are licensing matters.

Section 5.2.4 describes a standard site, Middletown, USA, in detail and examples of answers to the above questions are given in the description. In addition, Section 6.4 discusses the licensing problems associated with present day siting in the context of disturbances to the ecology: air pollu­tion, thermal effects, and aesthetic effects. The following section treats the effects of meteorology in more detail.

+ See also Section 6.4.

When the slug of sodium reaches the vessel plug, an impact results and the head will be damaged in two ways: the vessel plug may deform and lift against its restraints, and lateral forces after the initial impact may cause radial deformation of the vessel just below the head.

It is important to be able to compute whether any ingress of sodium is possible into the containment area above the vessel. Therefore the plug jump and radial deformation of the vessel head need to be computed from the sodium hammer impact. Codes do exist that perform this computation: The following equations illustrate a method of deriving the plug jump from a static load analysis

Momentum is conserved:

M3vs = MT i/p (5.25)

In these equations, M and v are mass and velocity, respectively, and the subscripts s and p refer to the sodium slug and the plug, respectively.

MT = Ms + Mp (including shields, etc.) (5.26)

Assuming that the plug restraint system, say bolts, is designed to hold the plug on while absorbing the energy of the sodium slug in the stretch of the bolts themselves, we can calculate this stretch knowing the number and size of the bolts. Bolts absorb energy,

Eb = average stress x elongation x volume (5.27)

and to absorb the total sodium slug energy

Eb = MTv* (5.28)

From this set of equations, knowing the initial sodium slug energy, the elongation of the bolts can be derived and thus the possibility of allowing a path for the ejection of sodium may be evaluated. Note that this calcula­tion is very pessimistic since some of the energy of Eq. (5.24) would actually be expended in the radial deformation of the vessel referred to above.

It is convenient first to outline the behavior of a poison that the fast reactor does not experience. In large thermal systems xenon produced from 135Te by way of 135I is now a very important poison. Because it is a byproduct of a fission product, it is flux-induced rather than temperature- induced. It is not an important poison in the fast reactor system.

The effect of xenon is complicated by its production from iodine and its destruction by a combination of neutron absorption and natural decay to 135Cs and by the various time delays involved. The production and destruc­tion is illustrated in Fig. 1.16.

n c

The relevant equations descriding the production and decay of iodine and production and destruction of xenon are:

dljdt = ахф — AjI dX/dt = AjI — AxX — ахфХ

After shut-down, the xenon concentration does not see the lack of pro­duction of tellurium for a time delay of several hours due to the 6.7-hr iodine decay time. However it does see an immediate reduction in removal by neutron capture; thus the xenon concentration grows. This increase in xenon concentration complicates the subsequent start-up of some small thermal reactors due to the poison increase.

Further, xenon poisoning could produce spatial instabilities in the very large thermal systems, as some regions of the core could see poison changes on different time scales from others. This effect is corrected by regional control systems.

In the fast spectrum, xenon is not important since its parasitic absorption is very low (Table 1.4) and the fast cores are too small for spatial instabili­ties. However, see the thorium cycle reactivity change in the next section, as it is a similar though opposite effect.

Other fast reactor absorbers are given in Table 1.6.

Although control rod malfunctions cannot give rise to large step changes of reactivity, during any survey of acceptable reactivity changes, it is useful

to include a parametric survey of the effects of large step changes. As will be seen, it is possible to envelope many other reactivity effects within an acceptable step value.

Figures 2.17 and 2.18 show power rises and temperature rises for a range of step additions which show that approximately 600 would be acceptable for the particular control system considered. It is important to realize that this acceptable value is dependent on the failure criterion chosen (here it is incipient fuel melting) and on the response of the protective system.

Many of the other postulated mechanisms for adding reactivity that follow will actually result in total reactivity changes that are less than the accep­table step value. Thus all can be shown to be acceptable without performing separate reactivity transients in each separate case.

The previous sections have discussed failure criteria from the point of view of anticipated failure mechanisms arising from faults. We also need to know how fuel pins may survive under normal conditions and so cladding survival criteria can be defined in terms of the following.

(a) Once-off conditions due to operational transients which might result in: overstress due to internal loadings; overtemperatures due to high power- to-flow conditions; and defect failure which is exhibited as an aggravation of other modes of failure. These conditions all arise from minor accidents or transient overshoots such as those discussed in Chapter 2.

(b) Cyclic conditions due to repeated minor loadings: stress cycles; temperature cycles due to load changes; external loadings, possibly from flow disturbances; and vibrational fatigue from flow induced vibrations.

(c) Continuous adverse conditions due to normal operation: cladding erosion; cladding corrosion; and internal loadings due to fuel swelling and fission gas pressure.

In these cases the cladding would have a failure because of a reduction in allowable cladding strain before yield. This might, in combination, reduce the failure limits discussed in Section 3.1.1.

The calculation of the energy release typified by Eqs. (4.15)-(4.22) is complicated by the need to solve spatial as well as temporal derivatives in a finite difference manner. Thus computer solutions are needed, and the available codes reflect improvements in the analysis over the past two decades (see also the Appendix).

(a) AX-1 (13a). This is a one-dimensional spherical model with a linear equation of state. It accepted only a step of reactivity and had no Doppler feedback.

(b) АХ-TNT (13b). This development of AX-1 was also one-dimensional but it included a linear or Clausius-Clapeyron equation of state. It had partial Doppler feedback, and it accepted reactivity ramps as well as steps. It had a total of 320 mesh points and it used an additional routine to cal­culate the available work energy in terms of TNT equivalent.

(c) WEAK EXPLOSION (14). Still a one-dimensional spherical model code, it included an equation of state of the form p = В exp [—A/(E + Eq)] and a Doppler feedback which depended on E1/2. This code had no calcula­tion of the work energy available for damage to structure.

(d) MARS (15). This code is two dimensional (r, z). It allows in its modified form all types of equations of state, all forms of reactivity addi­tions and allows the input of tabulated power distributions. Six core regions involving a total of 380 mesh points each can be used to map the core. Doppler feedback is included. A modified version of MARS calculated the available work as an isentropic expansion of fuel following termination of the transient by the dispersion of the core. However, modifications to the core model have allowed the available work to be calculated as a function of the amount of sodium present in the core into which energy is deposited by the molten fuel. This sodium is then assumed to undergo isentropic expansion.

(e) VENUS (12b). This latest code is still under development at ANL. It combines a comprehensive hydrodynamic code that calculates the move­ment of fuel and structural material allowing a pointwise variation of density with time, with a point kinetic model of the kind used in MARS. The code is produced as a module in a much larger series of codes which starts with a dynamic module and ends with damage calculations within the vessel. It is the first to avoid the use of constant density, although the results do not differ greatly from MARS except in special cases.

EBR-I, shown in Fig. 4.38 in cutaway form, was a 1 MWt reactor built in 1948. It was composed of 0.5-in. diameter pins of enriched 235U in a single big assembly which formed the entire 8-in. diameter core. Cooling was provided by NaK at 6 ft/sec increasing in temperature from 250 to 350°C. The vessel and piping were all doubly contained, top inlet being provided

to the vessel, going down through the blanket, and up through the core. The coolant was pumped by EM pumps. Control was provided by bottom operated rods and reflectors for fast control with a movement of the 5 ton bottom blanket to provide slow control. This latter control suffered from poor clearances.

Having set the limits on any radioactive release which may be permitted from the plant and having designed the plant to be accident free and so to give rise to no emissions, nevertheless it is prudent to design the contain­ment to provide protection in the event of an accident. Thus a containment evaluation accident is chosen as the worst accident to which the plant could ever be subjected, and this accident is used to assess the adequacy of the containment design. In many cases there will be an iteration between the calculation of accident conditions and the design of the containment.

Figure 5.5 shows that for the particular containment concept chosen, a radioactive dose in excess of the limits set by 10 CFR 100 could be ex­perienced following a core disruptive accident or a sodium fire with high radioactive contamination in the sodium only if a large number of safety features had failed or were not provided. Such a fault tree demonstrates that, for radioactive limits to be exceeded, a large number of conditions must arise at the same time. This can be seen by the large number of INHIBIT and AND gates present.

The sodium fire analysis is treated in Chapter 4 and the core disruptive accident will be treated in full in Sections 5.4 and 5.5 of this chapter. It suffices here to say that the two accidents either alone or together could give pressures in the range of 5-35 psia within the containment volume and temperatures of 200-700°F on the inner surfaces of the containment walls. Under these conditions the containment building is designed to comply with the leakage rates required to meet the dose limitations at the site boundaries.

The regulatory processes following an application for a construction or an operating license involve judges, juries, witnesses, plaintiffs, and the respective counselors and it is important to get the interrelationship of the various bodies into perspective.

* Note: this notation indicates volume percent per day, abbreviated throughout this volume as vol%/day.

6.2.1 Bodies Concerned with Safety for Nuclear Power Plants in the USA

6.2.1.1 Atomic Energy Commission

The AEC is involved at many stages of plant design and licensing through its Division of Reactor Development and Technology and through its Division of Reactor Licensing and the Advisory Committee on Reactor Safeguards.

(a) Division of Reactor Development and Technology (RDT). This divi­sion deals with the technical side of fast reactor technology rather than with licensing. It places contracts for all aspects of safety research with industry and the National Laboratories. It is concerned with the promotion of Nuclear Power. It is also available as an expert witness during licensing.

(b) Division of Reactor Licensing (DRL). This division, which deals with standards, licensing and compliance, checks the safety submission for formal compliance with safety criteria.

(c) Advisory Committee on Reactor Safeguards (ACRS). This committee makes the prime decision on the safety of the nuclear power plant. It receives the presentation of the safety evaluation from the applicants and after advice from DRL makes recommendations to the AEC Commissioners.

Fig. 6.2. United States Atomic Energy Commission organization (1971).

in 1971. It has been suggested that in the future the licensing process (DRL) should be separated from the technical promotional side of the AEC work (RDT).

The ACRS is composed of a set of independent non-AEC experts from a very wide range of disciplines. In 1971 they included: a professor of nuclear engineering, a physicist at Brookhaven National Laboratory, a consultant in metallurgy at Battelle Memorial Institute, a consultant in mechanical reactor engineering, a professor of chemical engineering, a consultant in hydraulic engineering and lake biology, the chairman of the board of Crown Central Petroleum Corporation, a senior engineer at Ar — gonne National Laboratory, a consultant in industrial chemistry, a pro­fessor of civil engineering, a consultant in chemical engineering, a physicist at Los Alamos Scientific Laboratory, a director of the Sanitary Engineering Research and Radiological Research Laboratory, a professor of nuclear engineering, and a senior physicist at Argonne National Laboratory.

The chairman of the committee is changed annually from among the committee members. The 14 members encompass a wide range of talents, and it is clear that such a committee makeup is liable to produce a very balanced outlook of the overall safety of any given nuclear plant under consideration.

Within RDT there is a technical organization led by a director (Milton Shaw for the year 1971) and assistant directors in charge of reactor en­gineering, nuclear safety, project management, reactor technology, plant engineering, program analysis, engineering standards, and army reactors.

Within the nuclear safety department there are groups working on research and development, engineering and tests, analysis and evaluation, and environment and sanitary engineering, although almost all departments have interests within the safety field.

RDT in practice has a dual function. As the technical advisory arm of the AEC, the division will enter the licensing process as the AEC expert advisor and, therefore, the division must keep abreast of technical develop­ments in all the relevant fields. RDT also places contracts for safety research and development with national laboratories and industry and the division is therefore in a position to direct national research and development to a large extent as it administers the available funds. It bears great responsibility for the direction of the LMFBR program.

The transient behavior of the neutron flux in a reactor core is represented by the diffusion equation (5):

dnjdt = exp(— Д2т)Лтс(1 — /3)2а<£ + exp(— B2rD)p £ ).{С{ + DVty — Елф

1 (1.3)

which expresses the fact that the rate of increase of neutrons is equal to the prompt and delayed neutron production reduced by leakage and ab­sorption (5).

The delayed neutrons’ concentrations are represented by;

dCJdt = (к^^фір) -Xfii, і = 1, 2,…, N (1.4)

which shows the radioactive decay of the precursors of the delayed neutrons.

Equation (1.3) is space, time, and energy dependent [и(г, t, £)]. We could more accurately start with a set of equations each of which applies to a different energy. Here are diffusion equations for three discrete energy bands as an example:

Fast group (> 1 MeV). Here fast neutrons are born, there is no resonance capture, but leakage occurs and the neutrons scatter down in energy to the next group

dnF/dt = [£„,(! —- Р)£афт/р] + Df — Е$ф$ (1.5)

In the fast reactor 2F is small.

Epithermal group (1 keV-0.5 MeV). Here delayed neutrons are born, and further neutrons arrive from the fast group after energy degradation

N

dnFjdt = pEF exp(—B2r — Td])^ j + P ^ + DF V2<f>^ (1.6)

t-i

In the fast reactor 27E is even smaller than 27F.

Thermal group (0.025 eV). Here all neutrons arrive after further energy degradation in a thermal reactor, although very few neutrons would reach this range in the fast reactor core

dn-i/dt EF exp(—52Tj))^E И — p (1.7)

These three equations may be summed to give Eq. (1.3) by defining a new diffusion coefficient D as:

D = DT + [p exp(—BH)DF Р2ф?/Р2фт] + [exp(—B2rT>)DF Р2фъ/Р2фт]

(1.8)

n = nFp exp(—B2t) + nF exp(— BHj)) + nT

Thus the original single diffusion equation can be used to represent the single average group of neutrons. So long as D is chosen correctly the equation is more accurate than is, at first sight, apparent. Note that in a steady state, Eq. (1.3) reduces to the critical condition keg = 1.0.

This reduction to a single group equation would also apply in the case of a fast reactor; however, one should note that it is an approximation that may be used in kinetic calculations while the accuracy would certainly not be adequate for steady-state physics calculations.